Intro/Outro (00:03):

Welcome to Supply Chain. Now the voice of global supply chain supply chain now focuses on the best in the business for our worldwide audience, the people, the technologies, the best practices, and today’s critical issues, the challenges and opportunities. Stay tuned to hear from Those Making Global Business happen right here on supply chain now.

Scott Luton (00:32):

Hey, good morning, good afternoon, good evening, wherever you may be. Scott Luten and Vin Sista here with you on Supply Chain now today. Welcome to today’s live stream, vin, how you doing?

Vin Vashishta (00:42):

Good, thanks for having me on. How about you?

Scott Luton (00:44):

Doing wonderful, and we had such a great time. Last time you joined Vin and heard so much from the market. We had to book you through your agent. You had a quick departure from your World Rock and roll tour just to join us a little bit here today, Ben, so I can’t wait.

Vin Vashishta (00:59):

Happy to be here. Thanks again.

Scott Luton (01:01):

So what instrument do you play, by the way, on that rock and roll tour I’m talking about, is it guitar, drums, tambourine? What is it?

Vin Vashishta (01:09):

I play the data.

Scott Luton (01:12):

Oh, that’s a great place to be. Great place to be. Well, hey folks out there definitely is. Stay tuned for a great show. We got teed up. It’s the buzz where every Monday at 12 noon eastern time, we discuss a variety of news and developments across not just global supply chain but global business. Now, today we’re going to be talking a lot about technology, especially artificial intelligence, and we got the right guy to join us to do just that. So we’re going to talk about AI infused my mouse peripheral devices. How about that? That’s raising a few eyebrows to discussing just exactly what is AI watching these days and a whole lot more. And folks, we want to hear from you. So give us your take and the comments throughout the show. And hey, if you’re listening to the podcast replay, you ought to consider joining us live each week. Again, Mondays 12 noon Easter time on LinkedIn, YouTube X, whatever. Come join us. Van. One last thing. Those folks are going to want to do this. It’s going to be a great show We’d love for you listeners or viewers out there. If you’d like what we’re talking about here today, hey, share it with your friends or your network, they’ll be glad you did. Then sharing is caring and hope we could help our friends and family members and colleagues out there to do life a little bit better, shouldn’t we?

Vin Vashishta (02:20):

Oh, definitely. Spreading a little bit of knowledge is a good thing, especially on a Monday.

Scott Luton (02:25):

That is right Monday, indeed. Alright, so Vin, looking forward to a great show. And by the way, some folks may remember that you are out in Reno, Nevada, beautiful Reno, Nevada, and I think it’s one of the best times of year weather wise to be out in Reno. Is that right? Vin?

Vin Vashishta (02:41):

Oh yes. Spring out here in the Tre Meadows. Absolutely beautiful. Especially up at the lake. It is amazing. I mean, I wouldn’t advise going into the lake yet, but everything else about this is absolutely beautiful. Time for Reno.

Scott Luton (02:52):

Might be a little bit cold to go skinny dipping just yet. Alright, so let’s share a couple resources with everyone and then we want to start with our newsletter we dropped over the weekend with that said, so y’all check this out. Look at that gorgeous vehicle there. So in addition, number 77, we touched on a variety of things. The US Postal Service has a lot of folks asking, Hey, where’s my mail? We talked about what’s some of the leading digital transformers out there are focusing their tech investments on and not just AI folks. Well beyond it, of course the United Auto Workers scored a big win recently as the Volkswagen factory in Chattanooga, Tennessee voted to unionize. Speaking of the automotive industry, VIN got a question for you. One of the things we talked about in this newsletter of the weekend is where did all the two door cars go?

(03:36):

So in the shot right there, that’s a 1986 Buick Grand National, right? But then as I shared pre-show, what was inseparable from my childhood was this car right here, this is the Chevrolet, this is the mid eighties or so, Chevrolet Monte Carlo Ss. This was everywhere. I think every family member must have had this. Every friend must have had this vehicle. I thought it was a really cool thing to do. Of course, a fellow two-door car. But then what about you? When you think of those formative years, especially the cool eighties, what two-door car was inseparable from your upbringing?

Vin Vashishta (04:09):

Well, not quite the eighties, but my first car in the nineties was a Nissan 300 zx. It was a two door and it was one of the first years they made one of those. And oh my goodness, it was so beautiful.

Scott Luton (04:23):

It was and fast, right? It’s pretty fast.

Vin Vashishta (04:25):

Well, yes, yes, it was. It fast to start up, but let’s just say the top speed not so great. But yes, the acceleration was wonderful.

Scott Luton (04:35):

Oh hey y’all, check it out. Let us know what you think. What two door, four cars come to mind because they’ve really, as the article points out, they’ve disappeared. In fact, cars have largely disappeared according to, I think it was Cox Automotive, I believe it was only 20% of new cars sold these days are cars. Most of ’em are SUVs or crossovers or whatever else. So y’all check out with that said, we dropped a link there in the chat. You one click away from checking that out. And let’s see here. Alison says, great to see you Alison. My husband believes that no sports car should never make with four. Hey, I got to say I don’t disagree, Ben, what you think?

Vin Vashishta (05:13):

I mean the Panama Porsche proved that it is possible. And Mercedes has all of its a MG lines. There is a hatchback, A MG. The E-Class has a hatchback.

Scott Luton (05:23):

Really,

Vin Vashishta (05:24):

Man. Yeah, it looks like a family car flies like a sports car.

Scott Luton (05:27):

Okay, well so Alison, you’ll have to report back, ask Matt what his favorite two-door car is and Alison, what is one of your favorite four-door sports cars, right? But great to see you. Looking forward to collaborating really, really soon. Alright, so Vin, a lot of good stuff. I want to share one other thing. This was a scene from a little video I put out there last Friday as part of my good news message and it’s two of my kids on a gorgeous day back in February. And it dawned on me that these little moments that make up our lives, especially as fathers or husbands or just family members, man, it’s some of the best times. It’s the simple things, right? So last Friday though, I spoke about how business leaders got to lean into this massive opportunity that exists right now, do things different to be able to empower your team members and give them this valuable time back so they can do more of stuff like this and hang out with their family and de-stress, take some pressure off. It’s really important that business leaders lean into that right now. Ben, would you agree?

Vin Vashishta (06:30):

I absolutely would. I can’t tell you the number of tech executives that have to explain. There are studies that show you break 45, 50 hours a week for too many weeks, productivity just falls apart. It’s time like that. That makes people smarter, more productive,

Scott Luton (06:43):

Bad things happen and it builds and it builds and it builds. That’s a great comment there, VIN. And there’s lots of studies out there. There’s a ton of pressure on workforces around the globe, whether you’re in supply chain or technology or elsewhere. So we’ve got to find ways of doing business better and much different. Alright folks invite you to check out that good news we’re just talking about. They got one link right there, one click away from checking that out and commenting and give us your take as well. Alright, VIN, one more announcement before we get to three stories plus a bonus meme today with the one only Vin Basta. Y’all check this out. Coming up soon. Well, you know what? It’s a Monday. I did not include my graphic for National supply chain day. So folks, you’ll have to take my word for it coming up April 29th, it is time to shine a big old spotlight on the individuals propelling our global supply chain industry forward propelling all of us forward.

(07:34):

Whether you’re operating machines out on the factory floor or demand planning or sourcing new suppliers or driving trucks coast to coast around the world, no matter how you contribute to making global supply chain happen, your story deserves to be spotlighted and celebrated. And we’re doing just that as we reinvigorate national supply chain day on April 29th. Folks, if you want to learn more, you can either click on this handy dandy link that our team just dropped there, so check that out there. And you can also go to supply chain now.com/nscd. If you’re listening to the replay, check us out and join us on April 29th at 12 noon eastern time while we celebrate an incredible, incredible industry. Alright, speaking of an incredible event, we got a ton of really cool stories to get to here today and I can’t wait to get to hear you weigh in with your take. So before we do that, some folks may have missed your last appearance with us because we blink and it’s been quite some time, maybe close to a year. So I want to make sure folks, for context, some of our listeners may not know your wealth of experience and expertise in industry. Your business is blown up left and right as y’all continue to grow and move mountains out there. And you’ve got a ton of that experience by the truckload in the technology space, right? Yes.

Vin Vashishta (08:46):

Almost 30 years. Wow. In technology founded V squared in 2012, it’s one of the oldest data and AI strategy r and d really just focused on that area of technology companies in the world. Been fortunate now to start a second business which handles training called high ROI certifications. Now we’re trying to train and upskill technical professionals on the business, teach businesses how to do data, ai, all of those pieces. So they actually make cash with it, man. And it’s been, I got to tell you, it’s been a great ride, a lot of fun, especially last few years, obviously

Scott Luton (09:20):

The golden age and you were doing ai, machine learning, all these technologies before it became cool and incredibly powerful as much as it’s today, right?

Vin Vashishta (09:31):

Oh yeah. We didn’t know what to call ourselves when I first started. We were making up names. Well

Scott Luton (09:36):

Folks, we’re going to hear from a real expert here today as we work through these stories and we definitely want to make sure that you’re able to connect with Vin who knows how y’all can collaborate. We’ll touch on that before we wrap here today. Alright, so getting into our first story, I’ve got this graphic, I promise Vin, I got this one. Alright, so from our friends over at the Verge, we’re talking smarter peripheral devices. That term brings me back to computer class in 1989. But hey, I bet a lot of our listeners are familiar with the company Logitech as they make a wide range of technology equipment. I think four of my 12 mice might be Logitech devices. The company has recently though rolled out a new mouse that will have a dedicated button to launch chat GPT. And right now, on that note, it will only work with chat gp, not any other AI chatbots.

(10:22):

More to come on that. Its goal seems to be making it easier for users right there in a moment to leverage AI to improve their work. Whether they’re writing something, figuring something out, tracking recipe ingredients, you name it. But if you don’t want to spend the 50 bucks or so to buy the new mouse with that dedicated AI power button, like power wheels or something that launches this prompt builder, well you might still be able to map that offering this new offering to your current mouse if it’s relatively new and be on your way. So Vin, I want to ask you your thoughts here. Is this a big deal or is this just more press for all the wrong reasons?

Vin Vashishta (10:56):

Yeah, I think we put AI in a little too much. Not everything needs to be an AI product, right? And there’s the challenge is a company like Logitech is looking at being disrupted almost entirely. What generative AI and more of what we’re rolling out year by year does is it replaces things like keyboards and mice. We already have touch displays. If you have a cell phone, you don’t use a mouse. If you have a tablet, you’re probably not using a mouse either. So Logitech is really challenged by number one, touchscreens, but also number two, ai. If we’re talking to our pc, your keyboard, it’s gone, your mouse, it’s gone. Your webcam is really, you could see their product portfolio is shrinking rapidly. This is their attempt. And I mean, I get it. I’m not singling Logitech out as doing something wrong, it’s their attempt to handle that, that sort of disruption that’s coming their way because generative AI really, it’s the new keyboard, it’s the new mouse, but I don’t know if this is the way should be handling it. I don’t know if this adds a lot of value and that’s sort of the problem.

Scott Luton (11:58):

It is the problem, right? Because you can cause a lot of harm with this AI washing, which we’re going to touch on in a second, but also as a consumer and vin, this is your space, so you have a leg up on many folks, but I’ll proudly call myself a non technologist, right? I don’t do any coding, I don’t even build AI platforms or anything. So when I come across the millionth latest device that touts this ai, but it doesn’t really offer a practical and powerful avenue of tapping into ai. It’s just more heartburn, right? And there’s some brand risk there as well, Ben, that you’re kind of pointing to as well, would you say?

Vin Vashishta (12:34):

Definitely is. And I think that the challenge is, number one, we want to use AI to drive new revenue. We want to put features into our products that could drive revenue. But what logitech’s really done is with one button press, you can launch AI or you can just go into the bottom right hand corner if you have a Windows PC and launch ai, I don’t need the extra button.

Scott Luton (12:57):

Features and benefits. Features and benefits folks, we’ve dropped the link to this article from The Verge. You can check it out as well. And hey, if you splurge on this new device with the power button, let us know. Let us know how it hopefully changes your day or makes your day easier. Alright, so before I move to this next related story on AI washing, I think this is a fascinating topic. I got to call timeout because folks, as y’all can see just over vin’s left shoulder, it is a bunch of baseballs from major league baseball games. Now we uncovered or we uncovered then that you’re a big baseball fan, but you’re a big San Francisco Giants fan, right?

Vin Vashishta (13:34):

Yes.

Scott Luton (13:35):

Some of our listeners and viewers out there may remember, if you’re baseball fans, my Atlanta Braves and his Giants spent years in the NL West and had some Duke it Out seasons came down to last days in some of those cases. So then little sidebar, what’s one of your all time favorite Giants players?

Vin Vashishta (13:55):

Oh, Barry Bonds. Easy. Buster Posey has got to be mentioned at the same time. Those two, one and two, some of the greatest that our team has had, I’ve got to say, Barry Bonds not getting into the hall, just absolute robbery. That’s heartbreaking. I understand the PED thing, but it was so common back then it was just heartbreaking to see it.

Scott Luton (14:12):

Well, I’ll tell you, both those players were incredible. Buster Posey killed the brave throw at his career, that Hall of Fame catcher. So while we’re talking Hall of Fame, you start a campaign to get Barry Bonds in and I will continue my campaign to get the Murph Dale Murphy in the Hall of Fame. Talk about crying shames too then. But anyway, so can we get a twofer if Barry gets in? Dale gets in. Sound good to you?

Vin Vashishta (14:34):

It sounds great. Unfortunately, I think Barry’s out of ballots. I think he’s already gone through the number of ballots that he can be on and be passed up on. So I think it is over for Barry Bonds, which is, like I said, it’s hard to have seen that.

Scott Luton (14:48):

Alright, so Mr. MLB Commissioner, I can’t remember your name right this second, but hey, give us a call, give me Vin a call. We’ll get it all squared away. We’re trying to solve the world’s problems. Vin, come on. Alright, moving back to AI in today’s buzz. I love this graphic here. And we’re talking about AI washing, right? Yeah. This is a great read from Tech Target that dials in on this notion of AI washing. So let me summarize it. And Ben, I can’t wait to get you to weigh in here. So in a nutshell, Ben La Kevi, I think I said that right? The author of this article points out that everything has claimed to be ai. Well, a lot of it isn’t just what we’re talking about a minute ago. And before you go and think, well, that’s just marketing, it’s harmless. No one gets hurt.

(15:29):

Well, hey, consider what Gary Gensler, chair of the Securities and Exchange Commission said a couple months ago. He warned in a speech at Yale University that companies that participate in AI washing, well, you could be breaking US securities laws. That’s not fun. You could be harming in investors, which is serious, very, very serious. And of course what we were talking about a minute ago, also misleading consumers, which has got a lot of brand risk to you do it your own peril. So then weigh in on this notion of AI Washington. What are your thoughts? I

Vin Vashishta (15:58):

Think the biggest danger is with something similar that happened with Devin, which is supposed to be a replacements for software engineers. There was a demo that was done where somebody went on social media and basically said, Hey, give me projects for web development and I’ll do them. And they claimed that they used Devon and were able to do all of these projects. And when someone dug into it, researcher dug into it, found out it wasn’t real, was not actually that capable. Now the people that are building Devon, they’re trying to do a good thing, but the challenge comes when c-level leaders see this and they say, oh, this can replace all of our developers. Oh, this is really ready to go. So we’re going to start laying developers off. And it’s not until six months, a year later when a lot of these projects start struggling in production, when customers start complaining or internal users say, this isn’t what we need, this isn’t working, you begin to see that there have been workers displaced, but quality of work and the business outcomes begin to suffer.

(17:00):

So that’s the biggest danger in my opinion, is that we continue this, it’s great. It does everything. It’s in everything. And when we overemphasize the AI and we sort of hide the fact that it’s not reliable enough to work for these particular use cases or doesn’t really cover as much as a person could and the person is still critical to be part of the loop. When we overlook that, it’s really the perception that begins to drive business decisions. And those business decisions can have some negative impacts. So that’s the important thing to take away from this is it’s very difficult to evaluate AI products and you’re not just evaluating functionality, you’re evaluating reliability. Just because it gives you an answer doesn’t mean it gives you the right answer. Just because it spits out some code doesn’t mean it spit out the right code, doesn’t mean that it’ll run. So there is this sort of chasm between functional and functional reliability. And if we don’t evaluate both sides of the coin, you can end up believing a whole lot of this whitewashing where the hype is much bigger than the actual product that gets delivered.

Scott Luton (18:12):

Vin, what timely expertise and perspective there folks, I at a very high level summarized article, you need to go read this article from the Verge and we got the link right there. You can check it out. They go into a lot of different examples and whatnot. vin, based on what you’re saying there, I got one quick follow-up question. I hope this isn’t a dumb question, but I love your emphasis on the reliability part and kind of what you’re implying about accurate outcomes that you can run a business based on, right? As folks out there are evaluating AI products, right? And that’s what they’re being told and fed, what’s a question or two to arm folks with that will help them kind of cut through the noise, cut through the marketing speak and really determine whether or not it’s truly the right outcome producing AI product or service? Any thoughts there? Ben?

Vin Vashishta (18:56):

You want benchmarks that have nothing to do with the model. You want benchmarks that have to do with outcomes you care about, and that’s the most important thing. You’ll hear startups launch these models or launch their services and they’ll say, oh, it’s more reliable than, or it’s more accurate than none of that matters because accuracy at the model level doesn’t necessarily translate into functional or reliability differences. And so this is the big question you’ll hear larger software vendors and more business-centric startups explain this is the productivity improvement that we’ve seen from businesses that we’ve worked with. Everything that you’re considering buying should have gone through and what’s called an alpha release, which is that first iteration, we’ll send these products out to a company and these are our best clients, our best customers, and they’re going to beat it up, they’re going to use it, they’re going to tell us this doesn’t work.

(19:48):

This does. And it’s sort of a shakeout period. And what we’re able to do during that time is get rid of all of our bigger problems that we didn’t anticipate functionally make it work exactly the way that our customers are going to need it to. But we also get those early results. We start benchmarking it to actual customer metrics, things that you care about. So don’t worry about getting into the technology terms. Ask what does this do and list your main KPIs, the biggest things that you are concerned with, the biggest strategic goals or outcomes, whatever it is that you are measured by. Ask them how they impact those metrics. And if they don’t have anything to give you, huge red flag means they don’t understand enterprises or potentially they just don’t understand your business because the product doesn’t target your main objectives. Yeah,

Scott Luton (20:42):

Well said vin. A lot of great advice there. And going back to your first point, don’t buy based on the model they’re using and what it does within that model, make sure you’re kicking the tires on what it’s meant to do specifically in a customizable fashion, unique fashion for your business. And another thing, I hadn’t really thought about VIN until you brought it up, of course only thing we talked about this last time. There’s a lot of folks out there that’s fearful of technology replacing their job. But the spin and the nuance you put on that example is when leaders make really bad purchasing decisions, misled, misguided purchasing decisions, and they think that they can save a bunch of labor costs with doing so and so really they lose twice. There are people that they unfortunately let go, that’s awful. And then the business of course doesn’t get what it needs. So it’s like a double whammy then.

Vin Vashishta (21:32):

It definitely is, and that’s the biggest challenge that I’m trying to address from inside of our industry where we need to be better and more accountable. As data scientists, AI researchers, we have to call out other parts of our industry and say, look, that’s just not possible. This isn’t feasible, that doesn’t work. And we have to call out some of these ridiculous claims because the hype is one thing. I mean, it’s great to be enthusiastic, that’s great, but the hype is something that we have to tamp down. We have to focus on real world impacts. And what I think really drives this is the disconnect between what really gets huge engagement on social media and what moves the needle for businesses. A 12% improvement in a specific area for a business could be a huge deal, but on social media, talking about 12% doesn’t really do anything.

(22:23):

You’re not getting all the likes and engagement that you want. You have to say 10 x, a hundred x, a thousand x. You have to say these crazy things in order to get likes and we have to start parsing them out. We have to say, look, this is what it can really do. And in almost every case, humans are still the center not engaged. They’re the center of all of the functionality and AI augments people. It makes us more productive. It allows us to do things that we couldn’t do in the past. But again, it allows people to do things we couldn’t do in the past. Will you need fewer resources in the future? Definitely. But you should think of that in terms of scaling. That means as your business gets bigger, you don’t need to hire more headcount. You don’t need to expand in that same old expensive way. You can use the same staff you have now to deliver a whole lot more to your customers. That’s the right way to think about it. Love

Scott Luton (23:20):

That vin doing business differently, leaning into the powerful new art of the possible out there the right way. Okay, love that perspective and I got to flash this image back on there. How long do you think, this isn’t really a serious question, how long do you think we’re walking the halls of businesses and we’re seeing bots take a quick lunch break like this vin huh? Around the corner next week?

Vin Vashishta (23:40):

It’s funny, we’ve already seen robots like this in companies like Walmart and Amazon and the larger distribution centers and hubs have robots obviously don’t look like that, but we do have those running around businesses and that’s what I don’t think we realize is exactly. I mean when you stop and take inventory of how much technology we have running businesses, we are getting to a point quickly where we are more dependent on technology than any other factor for running businesses and generating revenue. Yeah,

Scott Luton (24:12):

It’s unbelievable. Kidding aside, I think one of the most fascinating aspects of this age of technology we’re in, and every time I read an article related to this, it sends a chill up. My spine is sentience. And I think I said that right. What point in this journey are we going to get to a point where these bots and these technologies and AI can start to feel things? I know it might sound a little bit crazy, but last time you were here we talked about that movie her with Joaquin Phoenix. It’s from 2013 and it really folks, if you hadn’t checked it out, if you’re a sci-fi fan like I am, check that movie out because the ending will really have your mind racing. But then what’s your take on wind machines and technology and AI and chatbots, you name it, will be more sentient maybe in a scary way. I don’t know.

Vin Vashishta (24:59):

I don’t think we’re going to get there for a very long time because there’s a huge hurdle, and I’m not going to get too technical or too philosophical with it. Thank you, thank you. But there is a huge hurdle that we have to go from any of the approaches that we have now or that are on the near term horizon to something that would fit the description of sentient no matter what description we’re talking about. Because that’s the other thing, there are 50, a hundred, I mean you ask a hundred people, you might get a hundred different definitions. So that’s the near term, 10, 25 years having something that’s sentient. No, what we have to really think in terms of is that it doesn’t have to be really intelligence and capability are what we’re aiming for. We’re not aiming for consciousness.

(25:51):

So none of our approaches are going to, I mean unless something really accidental happens and in science, hey, it is a possibility. You do have unexpected, especially the more complex you get, the more unexpected behavior you get. But it’s really unlikely what we’re doing right now doesn’t even approach that. It’s complex, but we’re focused on capability, we’re focused on this construct of intelligence that isn’t a human construct. It is more of a capability construct. What can you do with it? How can you apply it? That’s what we’re targeting and it’s so far away from that sentient sort of conversation and capability set that I don’t see what we’re doing ever moving in that direction until a very long time in the future.

Scott Luton (26:35):

Okay, Ben, lemme ask you one quick follow up question, and I’m going off road a little bit here. By the way, T squared says, Hey, we need more Rudy’s and datas in the workplace. Of course data is a reference to Star Trek next generation. That’s an interesting show to watch. Ben, are you familiar with that reference? RUDI?

Vin Vashishta (26:51):

Nope.

Scott Luton (26:52):

Okay, so T Square?

Vin Vashishta (26:53):

No, I’m sure as soon as they say it I’m going to go. Oh yeah, yeah, that one.

Scott Luton (26:56):

That one. Hey, getting aside, lemme ask you one quick follow-up question, and I know this deserves a fuller conversation, but regulation, we’re in an age, not only a golden age of ai, but we’re seeing Europe and other geographic places across the globe enter into regulation conversations. Of course some have passed policies already. Ben, do you believe there’s a big need for regulation no matter where you live here across the globe?

Vin Vashishta (27:18):

I definitely think we need regulation, but we have to change the way we look at it right now, regulation is reactive. Something happens, we try to put a regulation in place and with every major technology development that we’ve had over the last 30 years, that’s a failed approach. What we have to do is get better people into each one of these committees, into each one of these government organizations that are responsible. Like in the US we have the FTC, we have, and now every federal agency now has a chief AI officer and we had chief data scientist, we have. So we have all of these different places, but the problem is that we don’t put the kind of resources and emphasis into these positions to get people who can get regulations in front of lawmakers that are in front of the technology not behind it.

(28:09):

Because once you’ve opened up the data genie, getting data privacy back is very, very difficult. Once you’ve opened up the AI genie, getting it to fall back into its box, very, very difficult. So we have to be in front, not behind. I think that’s the main thing that we need that’s different than the regulation we have. That’s why it feels like so much of our regulation doesn’t do what we think it should. That’s where a lot of the backlash to regulation comes from is when you’re reactive, oftentimes you put too much in place. Yes, oftentimes you’re trying to put a bandaid over something that’s become a massive problem, but that bandaid, it isn’t going to solve the problems that we have. Because a lot of times, especially if you think of major social media companies, your search engines, all of those, they have so much data about us already that you can’t put that genie back in the bottom. It’s already out there.

Scott Luton (29:00):

Well, appreciate you weighing in on that. I think we got to get it right folks in the room, right? And more folks in the room to your point, one of your first points, and man, just like it’s the age of ai, how we approach regulation, governance, security, all of that, we have a lot more work to do for sure, which is, I think that’s not necessarily a bad thing. It really speaks to this massive opportunity we have that we were talking about earlier. So a lot of good stuff there. I want to move right along to this note, this announcement from Big Blue, A-K-A-I-B-M who shared last week that met LAMA three is now available now vin, what’s the big deal here, you think? So

Vin Vashishta (29:35):

This is something that’s flying under the radar for most people that are not in the data field or the AI field. There are two camps that are competing with each other right now and it’s going to have a massive business implication going forward. You have closed models, closed AI where it is the traditional sort of walled garden. The company owns the model and most of what happens inside of the model is confidential. It’s proprietary On the other side, you have and IBM and Meta are leading a consortium in this direction. You have a open model approach where anyone can come out and grab Llama three or hugging Face is the main ecosystem for all of this, but hugging face makes almost a million different models available and you can just pull them down and start using them. That means they’re more transparent, you have more control over it.

(30:31):

Whereas with the closed model, you have less control, less transparency. And what could happen is if we let that consolidation with the closed win is that everyone’s sort of beholden to a very small number of companies to get their AI to get these advanced capabilities. So these two sides are sort of squaring off and like I said, huge business implications. If again, you have sort of this monopoly that has a stranglehold on one of the most influential technologies for the next 10 years, versus every company in this open paradigm can go out and use sort of this research that’s been done, these models that have been made freely available to anyone, any business, any individual, any academic institution that wants to use them. We’re also going to progress faster if we go with the open versus the entirely closed model. And so that’s the big dispute and if we leave it open, it is more accessible and you have more of this dynamism because the knowledge can be brought into companies easier and you don’t have to hire all the really expensive talent you would need to build these things yourself.

Scott Luton (31:41):

So Vin, lemme ask you, between the open and closed, right, this big way of looking at how industry is progressing, these two big factions here from a cybersecurity standpoint, what’s your first thought that comes to mind when it comes to open versus closed and the cyber risk there? Well,

Vin Vashishta (31:57):

The risks from closed are you have a single point of vulnerability and if someone were to figure out an exploit for one of the closed models, they get access to a whole lot of infrastructure data. I mean that one sort of model, it’s got a huge target painted on it, but it also has the most capable teams possible securing it because it’s such a high value asset. On the other side, on the open end, the ROI of an attack is lower, but the attacks are simpler in some cases because each one of the variants on these models could potentially have malicious data in there. Again, not getting too technical, but could have vulnerabilities that are really different than traditional software vulnerabilities. And so it’s very difficult for most companies to figure out does this open source model have those vulnerabilities or not? But if you’re pulling from large companies, companies like Meta, IBM, SAPs in there as well, you have all of these open model ecosystems and hugging face has them.

(33:07):

There are tons of credible startups. I mean Salesforce has released a model, Databricks has released a model. There’s all of these in the open and closed camps, and so there’s very capable ones that are out there in the open and closed. Both of those are really good secure options. But then you get into sort of the models that are variations of things like Mistral, which is a good solid foundational model. But there’s all of these flavors and all of these variations and if you pull the wrong one down, very difficult to figure out what your threats are. And like I said, completely different. And so both sides have pros and cons and there isn’t a clear, this is better. The best thing that you can say though is stick with credible companies that can explain their security practices. Look for credible open models that are put forth by these core foundational institutions. People there’s 50 some partners in IBM and META’S consortium. And so those are credible sources for models. Hugging Face is a credible company, but not necessarily every model on hugging face is a credible model. And so it starts getting kind of murky at some point. Yes

Scott Luton (34:21):

It does. I mean just listening to you walk through all of that last couple of minutes. Alright, so here’s the deal we got to make. Besides getting Barry and Dale Murphy into the hall of Fame as the supply chain now teams, we continue to figure out our technology path moving forward. We got to be able to reach you on your World Rock and roll tour so you can advise this vin, is that a deal leaving here today? Oh yeah.

Vin Vashishta (34:40):

Okay. Definitely, definitely.

Scott Luton (34:42):

Alright, good stuff. So folks, don’t take our word for it. Check out the release that talks about IBM’s announcement and Meta LAMA three, those open models. So check that out. You want to click away? Korah the one and only is with us here today, Korah, I look forward to seeing you in Gartner at a couple weeks. He says the biggest gap in supply chain is still the, so what does AI do really for us? AKA Rudy rule, distillation. Okay, it’s no one for me. I

Vin Vashishta (35:09):

Was trying to figure out what movie he was referencing.

Scott Luton (35:11):

Me. You both. Oh gosh. Alright, so Rah continues AKA Rudy, that rule distillation distills the knowledge of black box teacher models into rule-based student models. Statistics are composed into logical rules or into supply chain context. So we limit AI hallucinations. Good stuff there. As always. Korah, Jose, Ben, any comments on what Korah shares here? I mean

Vin Vashishta (35:37):

Outside of sort of embarrassing that I completely missed that one. I think what’s most important to figure out is that hallucinations are an inevitable component of not only large language models, but even simpler models. We don’t call them hallucinations, but they do have places where they will not provide accurate inference. And each approach has different ways of mitigating those challenges. And as you get a more complex model, you need more complex approaches. But the core tenet of everything that you can do to make your models more explainable to sort of remove that black box starts with data. Yes. If you’re curating data in such a way that you have context about what it is that generated the data in the first place, an assembly line, is that a truck? Is that some component of your business? Is that a workflow? Is that a process? Is that something customers are doing with your products? All of that context helps you build data to train models, but that data also makes the model more transparent. So start with the data and when you use that well curated dataset to train your model, your model is more transparent, you understand it better. This is where we start getting into some things like knowledge graphs and ontologies and some fairly technical constructs. But remember the term knowledge graph. You’ll hear it again if you’re in supply chain, you probably already have

Scott Luton (37:03):

No doubt by the way T Square comes back, rah was able to fit that in. So Rudy is from the Jetsons. That’s where Rudy is from. I love how Korah took that and baked it into his commentary. And Rah says, stronger in community, always. This is so insightful. I have to pay close attention to every word spoken most dense supply chain now today. Hey, it’s chock full. We warned everybody then brings it and Kora says, kudos to the team. Vince is an AI rockstar. Hey Korah, I agree with you. He doesn’t play the drums, he doesn’t play the rock guitar. He plays the data and the technology and that’s why he is part of the Rolling Stones of global technology. Korah, always a pleasure. Great to have you here today. Alright, so this last thing we’re going to ask you about, we’re going to talk about your book in a minute and make sure folks know how to connect with you, VIN, but I love your sense of humor, VIN, I love it.

(37:54):

And folks, if you’re not following or connected to VIN on social, you are missing out. Look at this next crap here. So VIN took this long time mean and put a little spin on it a few weeks ago. Now it might be a bunch of chatbots that are responsible for all the Facebook fighting. Now this statement, but check this out. So the meme is fighting the notion is thought provoking. And as is this statement that you led with, I think on this LinkedIn post that we stole this from, you said quote, the internet and social media have created a lot of space for mediocrity to thrive with so much digital space to fill standards have plummeted. AI unchecked will amplify the problem. Alright, VIN, tell us more of what you see out there.

Vin Vashishta (38:40):

Well I think the biggest challenge that we have is we used to have enough space for original thought. And so we used to prioritize the highest quality content to the highest quality ideas. Now with social media and just content in general, we have this massive amount of space to fill as a result. You have people who are really writing things that are filling space. You have a ton of room for bots. There was a recent study done that says about half of all of the traffic on the internet is created by bots or AI agents of some sort. That’s increased by 30% I think they said over the last five years. And so what’s going to happen is AI is going to take all of this bot content and low quality content and start learning from it. That means AI number one, it’s really good at space filler type content creation today.

(39:33):

So it’s going to put more filler content out there. And as a result of learning from a lot of filler content, you’re going to have this reinforcement cycle where top tier models won’t suffer from this so much, but other models will begin to amplify. They’ll basically be eating a lot of what is put out there, what’s produced by other models and becoming more like other models. And this sort of mediocrity cycle will begin to amplify and amplify and amplify. And so what you’ll see on social media, what you’ll see as far as content is concerned is the opportunity really is trust and original high value, high quality content thinking. So this is an opportunity for any business that has genuine thought leadership to position itself in a tier that’s so much higher than right now about half of the content, but really in that top 5% just by having true insights, real thought leadership, being able to entertain people in a way that you can also sort of inform them on the backend, which is my approach to it. And so this is really the opportunity for businesses look for ways that you can deliver insights because that’s going to build trust and trust is a new currency. Yes.

Scott Luton (40:49):

Well said then with trust is how we move mountains and we bring ecosystems together and strengthen the bonds there and the trade-off points. So I’ll love your final comment there. Trust is the new currency for sure. Alright, so I bet trust has always been important. I would argue maybe, and I bet you agree that now because of how fast we’re forming partnerships and making decisions and the overall speed of business, it puts your trust giving and receiving processes a lot of pressure on them. You don’t have three weeks as a contract gets routed back through the mail and all the different departments. I mean things are taking place in three seconds. And so it has really challenged our ability to really build trust the right way. I would argue. Vin, would you agree with that? Generally?

Vin Vashishta (41:36):

Definitely. I think that compressed timelines, number one, we are going to need more automation to figure out can we trust the contract? Is there an anomaly hidden in all of this? Can we parse this content to figure out what’s really inside of this proposal or what can this company truly do from a capability standpoint? So all of those are number one, we’re going to use sort of technology to figure out if we should trust, see the evidence of the capabilities or the promises that are being made. But on the other side of it, yeah, it really is just trusting and track records of success and being able to explain why we’ve been successful for customers in the past. I think that’s so critical. It’s not just what have we delivered, but why is it that we’re successful? If you know that you can repeat success. Well

Scott Luton (42:24):

Said. Well said, man. What an outstanding conversation here today, vin. And by the way folks, before we leave that last article, check out vin’s LinkedIn post on chatbots when social media is just arguing with itself. I love this. All right, there’s so many different uses for that meme that Spiderman meme. All right, so y’all check that out. Be sure to follow and connect with VIN out there. I’m telling y’all it will make your weeks better because it’s humor with actionable insights. Alright, so question for you as we start to wrap here today. For all the many, many business leaders, but especially supply chain leaders out there, VIN, that are still trying to find the most successful path forward when it comes to leveraging AI in their organization. I know that we could talk for hours about this, but what is one piece of advice you would share with folks out there? They’re trying to figure out this AI tidal wave.

Vin Vashishta (43:15):

The first thing is never lose sight of your core business. It’s not the technology business. First, business objectives, first business strategy first, technology strategy. Second, your technical strategy should amplify your business strategy. It should align with it, it should make it better. You should be seeing results faster. You should be seeing better results. It should be creating opportunities for you that weren’t possible before. Don’t worry about the technology. Keep asking that same question. What can you do? Here are my strategic objectives. What can you do? Here are my strategic objectives and trust the people that bring you answers. Love that. Those are the people to trust and forget about the technology because that’ll iterate. Generative AI will be generative AI or generative AI 4.0 in 20 minutes. The technology doesn’t matter. What can it do for you? So if don’t lose sight of your core business core functions, the focus, the strategy, all of that is still way more important. And look for people who can help you do what you do today better and show you what you can do tomorrow that you couldn’t do today. Look for those people.

Scott Luton (44:25):

Love that advice folks. One of the things that he really brought home is keep banging the drum because sometimes you might get a half answer and that doesn’t cut it these days. So keep banging the drum. What will this do for me? Here’s my strategic objectives. What will it do for us? The one-two punch is what VIN is advising there. Good stuff. Alright, so just over your right shoulder is your popular book from Data to Profit. Now you know what? I loaned my copy of that out to a friend and it just dawned on me. I have not gotten that back yet. Then. So I don’t have my prop like I normally do. We’ll track him down. He’ll owe a big library overdue fee like that episode of Seinfeld, right? But from Data to Profit, it’s been a popular read making a lot of waves out there. What’s been one of your favorite pieces of feedback from the market related to the book? Well,

Vin Vashishta (45:12):

I think it’s two-sided. It’s interesting that technology professionals, it’s dragging them into the business. It’s forcing technology professionals to come to terms with, if it doesn’t drive ROI, if it doesn’t create something for a customer, if it doesn’t create even in a nonprofit, if it doesn’t create societal good, it doesn’t matter. We’re here for business objectives, organizational objectives, government objective, whoever we work with or for, and it’s forced them. So we think of, oh, everyone needs to become more data literate and AI literate, but technical teams need to become more business literate as well. But it’s also opened a lot of c-level leaders minds to, there’s more to this and it’s going to be longer. This is never stopping. The amount of technology we use today is the least amount of technology we will ever use. This is as easy as it gets.

Scott Luton (46:05):

Oh my gosh, that excites me and scares me to death all at the same time. Vin, love that. And I loved your comment there a lot amongst your good stuff there. How onus just isn’t on business leaders becoming more technology literate, which we talk about all the time. But the inverse of that, we’ve got to have our technologists out there continually become more business literate. That’s such a great point, VIN. Alright, so folks, y’all can get the book from Data to Profit wherever you get your books. I think we might have a link. Amanda and Catherine will drop out there as well. Try to make it as easy as possible for folks. By the way, big thanks to Catherine and Amanda for making it happen behind the scenes as always. Alright, so Vin, as you’re about to resume your rock and roll world tour once a more, how can folks track you down and connect with you? Vin Vata,

Vin Vashishta (46:50):

Easiest way, data science. Vin, thank you to the French wine industry for naming a domain after me. Appreciate that you find everything there or obviously collectively on social media, LinkedIn,

Scott Luton (47:02):

It’s just that easy. It is just that easy. And folks, as Amanda Drops here, you can check out vin’s book right there via Amazon. I promise you’ll not be. Well, you know what? If you’ve enjoyed the last hour, wait till you read the book. Good, good stuff. More business leaders, more perspective and expertise that we all need to put between our ears. Okay, then a real pleasure for you to be here with us today. I love your perspective. I love how you keep it the right level too. Throughout your response, you kept saying, Hey, not to get too technical. You’re very aware of folks like the non technologists like myself that are in the room and how can we communicate for all folks, right? So we’re all aligned and are all understanding where we are with the conversation. That’s a great approach then that you bring to the table. So really appreciate you being here with us today. Vin Basta,

Vin Vashishta (47:51):

I’m so grateful for you to have me. Thank you so much. Great

Scott Luton (47:54):

Conversation. Definitely. I look forward to having you back. Alright folks, connect with Vin, track him down. We told you how to do that. Check out the book, make sure you’re connected with Vin across social, especially on LinkedIn, where you see some of this great humor and actual insights that I’ve enjoyed so much. Big thanks again to Catherine and Amanda. All the folks that showed up here today would know we couldn’t hit everybody’s comment. But folks, whatever you do, take vin’s advice. Take at least one thing that he dropped here today. Put it into action. Your teams are just, they’re hungry for doing business in new, exciting, and more successful ways. Try to find how you can give them time back and take some pressure off. So with all of that said, deeds, not words, folks, and the actions we take. With all that said, on behalf the entire team here at Supply Chain now, Scott Luton challenging you to do good, to give forward and to be the change. We’ll see you next time, right back here at Supply Chain now. Thanks everybody.

Intro/Outro (48:45):

Thanks for being a part of our supply chain now, community. Check out all of our programming@supplychainnow.com and make sure you subscribe to Supply Chain now, anywhere you listen to podcasts. And follow us on Facebook, LinkedIn, Twitter, and Instagram. See you next time on Supply Chain. Now.